Recently, the Hong Kong Special Administrative Region (“HKSAR”) Government’s Intellectual Property Department (“IPD”) has concluded a two-month public consultation, seeking views from the public to address copyright concerns that have emerged from the recent advancements in artificial intelligence (“AI”).
Prioritising the challenges posed by AI is essential, especially in view of AI’s close intersection with a wide array of domains, including innovation and technology as well as creative fields. Utilising data that may feature copyrighted content, generative AI yields a variety of rich outputs, ranging from written text, images, audio to video. Therefore, the implications arising from this novel technology are far reaching and merit careful scrutiny.
As delineated in the IPD’s consultation paper, the current provisions of the Copyright Ordinance (“CO”) can adequately safeguard creations generated by AI. The present legal framework offers protection for original works in the fields of art, drama, literature, and music, along with audio recordings, films, television broadcasts, cable shows, and the typographical design of published materials.
- For literary, dramatic, musical and artistic (LDMA) works, the “computer-generated works” provisions apply to AI-generated works without human authors.
- For non-LDMA works (e.g. sound recordings, films), existing provisions apply regardless of whether these works are created by humans or computers.
Legal protection as provided for under the CO already extends to those identified as “computer generated”. The IPD deems this classification sufficiently broad to incorporate creations generated by emerging technologies, including pieces produced by computer independently of human input.
The consultation paper features a proposed exception aimed specifically at text and data mining. This refers to a process that employs computational methods to examine extensive datasets, potentially including works protected by copyright.
The pros and cons relating to the adoption of a text and data mining exception are considered in the consultation paper. Proponents of such an exception emphasise the importance of widening access to copyrighted works for the purpose of training AI technologies. On the other hand, criticism levelled against such an approach includes the likelihood of disrupting established market behaviours and apprehensions over its long-term viability.
